ClassWeightsの設定方法
Show older comments
深層学習を使用したセマンティック セグメンテーションhttps://jp.mathworks.com/help/releases/R2018a/vision/examples/semantic-segmentation-using-deep-learning.html
をもとに自分で用意したデータセットで解析を行ったところ以下のようなエラーが出ました.

以下にi_learningのコードを示します.
%ネットワークの作成
imageSize = [360 480 3];
numClasses = numel(classes);
lgraph = segnetLayers(imageSize,numClasses,'vgg16');
%クラスの重み付けを使用したクラスのバランス調整
imageFreq = tbl.PixelCount ./ tbl.ImagePixelCount;
classWeights = median(imageFreq) ./ imageFreq;
pxLayer = pixelClassificationLayer('Name','labels','ClassNames',tbl.Name,'ClassWeights',classWeights);
lgraph = removeLayers(lgraph,'pixelLabels');
lgraph = addLayers(lgraph, pxLayer);
lgraph = connectLayers(lgraph,'softmax','labels');
%学習オプションの選択
options = trainingOptions('sgdm', ...
'Momentum',0.9, ...
'InitialLearnRate',1e-3, ...
'L2Regularization',0.0005, ...
'MaxEpochs',100, ...
'MiniBatchSize',2, ...
'Shuffle','every-epoch', ...
'VerboseFrequency',2);
%データ拡張
augmenter = imageDataAugmenter('RandXReflection',true,...
'RandXTranslation',[-10 10],'RandYTranslation',[-10 10]);
%学習の開始
pximds = pixelLabelImageDatastore(imdsTrain,pxdsTrain,'DataAugmentation',augmenter);
net= trainNetwork(pximds,lgraph,options);
Accepted Answer
More Answers (0)
Categories
Find more on Deep Learning Toolbox in Help Center and File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!